mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-26 01:12:21 +00:00
Customers were confused about what input vs output tokens mean and how cache tokens relate to the total. Adds a reusable TokenMetricLabel with info tooltips on the Cost tab's Total/Input/Output/Cache Read/Cache Write cards, documenting that input already includes cache tokens (so the breakdown sums to total) and that output includes reasoning tokens. |
||
|---|---|---|
| .. | ||
| litellm-dashboard | ||
| Dockerfile | ||
| nginx.conf | ||